Data Analysis and Machine Learning
We develop and use a wide range of methods from the field of data science and strive to transform data into valuable information and transferable knowledge. In doing so, we extract new information from simulations and experiments, identify patterns, structure, and trends in microscopy data, and ultimately improve our understanding of why materials, processes, and systems work the way they do. To do this, we use a wide range of different methods, from classical statistical analysis to statistical machine learning to different variants of deep learning.
One focus of our research activity is on data mining and high-throughput/on-the-fly analysis of microscopy data, but we are equally enthusiastic when it comes to other data, for example for extending classical simulations with data-based models. In many cases, "injecting" physical knowledge into machine learning models is an important amplifier for the models we develop.
Prof. Dr. Stefan Sandfeld
Building TZA / Room D1.15
Electron Microscopy Simulation and Data Science Lab
Deep-learning assisted fast in situ 4D electron microscope imaging (FAST-EMI)
Rhine-Ruhr Center for Scientific Data Literacy (DKZ.2R)